What problem does it solve?
This skill provides supervisors with reliable tools to manage subordinate Anima agents' processes, models, task delegation, and state monitoring so teams remain orderly and recoverable when agents need maintenance or reconfiguration.
Core Features & Use Cases
- Process control: Suspend and resume direct-report agents by toggling enabled state and preventing automatic reconciliation-based restarts when appropriate.
- Model management: Change an agent's primary or background LLM model (heartbeat/cron model) in status.json and trigger restarts to apply changes.
- Operational actions: Restart subordinate processes via reconciliation flags, delegate tasks with queueing and DM notifications, and track delegated task progress through a task tracker.
- Organization visibility & auditing: Produce a tree dashboard showing process status, activity, current and pending tasks, perform liveness pings, read current_task and pending task files, and run time-bounded audits (1–30 days) for activity summaries and error/tool-use statistics.
- Safety & scope: Actions that stop or change models apply only to direct subordinates; grandparents must request changes to deeper descendants, and no skill action can operate on the caller itself.
Quick Start
Disable the subordinate named aoi for temporary suspension with reason "operational pause".
